Bottleneck flows of pedestrians and grains
Alexandre Nicolas (LPTMS, Université Paris-Sud)
Simple models from Statistical Physics can be surprisingly useful to study practical problems that lie outside the traditional realm of Physics. I will illustrate this statement with two distinct (yet related!) examples: granular flows through a constriction and competitive passages of pedestrians through a door.
To destroy clogs in constricted granular flows, vibrations are often applied to the hopper or silo, but clogs do not yield instantly. Instead, they exhibit an anomalously broad distribution of lifetimes. This elusive feature was recently captured within a model in which we likened the destruction of clogs to thermally activated escapes from a set of energy traps; the model also reproduces other salient experimental observations [1]. Turning to the second example, competitive pedestrian flows through a narrow door look very disordered, and yet they were found to present robust statistical features, such as anticorrelated time gaps between escapes and exponentially distributed bursts of escapes. On the basis of simple models, we have shown that in fact these features emerge generically in constricted flows [2].
- [1] Alexandre Nicolas, Ángel Garcimartín and Iker Zuriguel, « A trap model for clogging and unclogging in granular hopper flows » (2017), preprint arxiv:1711.04455
- [2] Alexandre Nicolas and Ioannis Touloupas, Origin of the correlations between exit times in pedestrian flows through a bottleneck, J. Stat. Mech.: Theor. Exp. 2018(1), 013402.